The culture of normal oral epithelial cells and immortalized oral epithelial cell lines has been published (for example see Xu et al., Methods in Cell Science 18:31-39, 1996 and Sacks, Cancer and Metastasis Review, 15:27-51, 1996). Such cultures or preferably dysplastic oral epithelial cells are being used to screen potential chemopreventive agents. Initial growth inhibition or other cytotoxicity assays with the test chemopreventive agent present are being used to select appropriate concentration ranges for use in the subsequent transformation-inhibition assays. The levels of induction of three endpoints associated with inhibiting, reversing, or retarding the cancer process in response to exposure to potential chemopreventive agents are being explored. Normal oral epithelial cells and immortalized oral epithelial cell lines (dysplastic oral epithelial cells) are being used to screen potential chemopreventive agents. The endpoints chosen are amenable to quantitation of response. At least three endpoints shall be measured for response to chemopreventive agents. The endpoints include: retardation of cell cycle, increased glutathione content, and induction of apoptosis, changes in ploidy, proliferative index (PCNA, BrdU), Ki-67, MiB-1, p53, KS-5, nuclear pleomorphism index, RAR-beta, TGF-beta, and micronuclei. Appropriate controls are being simultaneously tested. The ED50 (50% effective dose) are being determined by a method determined by a standard protocol.